A few years ago, when I was first getting into the world of DIY pedals and other fun electronic bits, the idea became planted in my brain, perhaps by the government or aliens (pick one or make up your own), that FETs (field-effect transistors) are like sooperdooper special mojo transistors.
I was totally in love with my brand new OCD around that time, and a close friend pointed out that it used MOSFET clipping (metal-oxide semiconductor field-effect transistor). At that time using transistors for clipping was big news, LEDs as clippers were new to me then.
I hadn’t really thought about FETs much until some time last year when I decided I wanted to do something with them, so I began work with a single JFET (junction field-effect transistor) gain stage.
This project, and the two that followed are a great example of things taking a different route to the one planned, and the joy of seeing where that route might take you.
I had decided to first work on a single gain stage JFET boost using a J201, which would go in an 1590A enclosure (one of the wee ones). I breadboarded one which sounded alright, so I planned the layout, chopped up some perfboard and got it in a small enclosure.
I tried it in front of and after my Big Muff. When I put it before the Big Muff it added gain, and not much else. It sounded pretty saturated. I found if it was placed after the Big Muff it added more character. I originally had it set up so when the gain was turned all the way down, it was at parity, and turning it up added more boost, cos why would you want a boost to make things quieter right!?
I passed this first version on to a friend. I had mentioned that it sounded best after another drive pedal so he tried it last in his chain, after a drive pedal, going into a mixing desk for a home recording set up. This is when something strange happened. When it was turned fully anti-clockwise (which should have been parity) no sound passed through, and when it was fully cranked it was barely beyond parity. harumph. I later found out that some mixing desks have significantly lower input impedance than guitar pedals and amps. If you’re clueless about impedance check out this blog post. So my best and only guess is that the mismatch of impedances was shunting most of the signal to ground and sucking all the volume out of the pedal. The thing is, this wasn’t happening with his other pedals when placed last in the chain, which meant that the output impedance of that original JFET boost must have been far, far too high. I didn’t understand too much about impedance then, but it was definitely a cue to learn more.
A couple of months later I got the chance to run the schematic past wiser eyes. So, here’s the original schematic with notes.
In developing the Wee Fettle Boost I learnt a few things.
JFETs work differently to BJTs (bipolar junction transistors, 2N5088, 2N3904, etc.), whereas BJTs need a pull down resistor on the input, JFETs don’t. It’s something that I need to read up on more, but basically, pull down resistors are placed either before the input cap, or after the output cap to eliminate switch pop. When you turn your pedal on, you get a small amount of DC that needs to go somewhere. The pulldown resistor gives it a direct path to ground. The reason this isn’t needed on a JFET is that DC can’t pass from drain to gate. I may be wrong about this BUT, the ZVEX Super Hard On (also a single JFET boost) doesn’t have a pulldown resistor on the input capacitor. The Alembic Stratoblaster (another single JFET boost) doesn’t even have an input capacitor. In some schematics it has a resistor from input to ground, in others it doesn’t. See below.
But we still need a pulldown resistor right? Yup. After checking a few other single JFET boosts, including the two above I did notice that the pulldown resistor comes after the output cap AND in every case it is a 100K resistor to ground. Lightbulbs and bells flashed and dinged in my brain. I suddenly realised that volume pots (A100K pots in pretty much every pedal ever) are doing exactly the same job! Looking at my original schematic, I had a 500K pot AND a 470K resistor, meaning the ouptput impedance was up to 970K!!! No wonder it was causing issues! What a berk.
My friend suggested putting the pot on the drain as the gain control. He also mentioned putting a trimpot on the source to get the biasing right. JFETs can be a little bit tricky, and finicky, and you need to tune them just right.
Next time, I’ll show you how this all worked out.